Africa RISING M&E Expert Meeting

5-7 September 2012
Large auditorium, ILRI Ethiopia, Addis Ababa
Back to the event agenda

---

Open Space

How are we going to interact and collaborate on data collection? (Ryan)

IFPRI will coordinate with three (to be named) regional coordinators - one for each megasite.
In West Africa this will happen in November, in ESA in October.
The coordinators will work with people on the ground and ensure the data goes back up to IFPRI and to Spatial Dev.

Same set up for indicator forms:

Project and activity locations will be tracked by Steve.
--> everyone can give him the locations of all quick wins.
Sub-grantees will provide all coordinates (geotags) of all sites for all activities.

PMT will handle multiple locations for single activities.
Once quick win activities are entered, the rest will be much easier.

Spatial Dev will set up an export function to USAid.
Spatial Dev is planning derivatives (hi res satellite imagery), planning to use GIS data from the sites and planning to set up a data sharing site/mechanism (e,.g. FTP) so all GIS people can access that.

Collaborating on evaluation design

• We need to come up with data collection tools e.g. forms and checklists, focus group discussions, meetings, farmer field days, informal surveys with traders, extension agents and researchers.
  • Set up guidelines: IFPRI - IITA - ILRI
  • Implementation: researchers and partners
  • Timelines and budget need to be clarified
• Tracking price information (input/output) at village level - on rainfall / temperatures.
• Formal surveys e.g. baselines
  • Farm/household questionnaires
  • Community questionnaires
  • Other VC actors (seeds, fertilizers, agro dealers, processors, farmers, NGOs)
  • Mid-line repeat baseline
  • End line repeat baseline
• Set up guidelines - by IFPRI and national statistical departments (timing, budgets... emphasize EA surveys)
  • Draft questionnaires (by researchers in IFPRI and other partners)
  • Pre test (population, sampling)
  • Recruit enumerators
  • Supervise data collection and entry
  • Centralisation / access to users
• Experimental data:
  • Data protocols
  • Monitoring data (sampling)
  • Results
  • Farmers' feedback evaluation (for mother/baby trials)
    • Beginning
    • Mid-season
    • End season
• Secondary data
  • Adm data on area:
    • Area
    • Yield
    • Prices
    • Cost of production based on extension recommendations
    • Regulations / local policies
    • Demographics
• On prices:
  • Data collection / organization in the countries: MoA / Statistical departement and field staff - contract to collect data in target disctricts
  • Budget supervision by IFPRI
• Rainfall / temp - extension officers and farmers in villages
  • Supply rain gauge
  • Recruit
  • Training
  • Pay
  • Supervise
  • Researchers / teams - IITA-ILRI

Comments:

• Q: What do you mean by 'setting up guidelines' - is it to design questionnaires?
• A: Yes
• Q: Guidelines for data collection are for what?
• A: These are the same as mentioned before. We should have a structure to collect data from focus group discussions. All this is qualitative. Even for field trials we should do this.
• IFPRI should provide a template for indicators, in the same fashion as for design questionnaires.
• Q: Physically, are you going to collect data: electronically? On paper form, as text messages? Digitally, you can send it digitally to e.g. SpatialDev etc. and you will minimize errors in data transcription particularly in GPS transcription. Perhaps use some sort of tablets or PDA for data collection to send it digitally.
• A: Practically, we intend to just print forms and train people on data collection and collect these once per month. Collect all forms... --> We were thinking of paper forms because the investment to go for electronic forms is high. For GIS we just need the right devices. --> For HH surveys, it can save time and errors. // (Carlo) I was working on this before and I was planning to partner with a company working on electronic surveys etc. and they are currently running these surveys in Ethiopia etc. We will explore this. --> (Sibiry) The investment is going to be returned very easily. A tablet comes up at USD 200. For WA and ET this can be done - since unlike ESA there is more time to prepare this.
• Q: What is the plan for rainfall data? It's the most important factor to understand potential adoption of technologies and usually it can't be analysed well enough.
• A: What we plan: Each of the mother trials will have ?? at the sites. 30 - 40 mother trials to eb ranked. We'll trained lead farmers to monitor farmers who are doing that and we'll get very valuable data and we'll be able to do spatial and temporal analysis of rainfall. --> Is this a plan for all Africa RISING sites?
• The daily rainfall data should be available on all systems.

Timing of activities (Mateete)

Based on cropping systems: In Tanzania we need to plant by November.

1. Baseline surveys

• ESA: 1 Nov. 2012
• WA: Jan. 2013
• ET: Jan 2013

2. Appointment of M&E officers

• ESA: 1 October 2012
• WA: 1 Nov 2012
• ET: 1 Nov 2012

3. Field activities

• ESA: 1 Nov. 2012
• WA: March 2013
• ET: March 2013

Critical: site selection in Tanzania.

Comments:

• In Ethiopia and West Africa we'll need to start earlier than March 2013.
• Have Nafaka narrowed down villages etc.?
• (Carlo): we're planning to have the M&E officers asap but we have difficulties recruiting one coordinator in Washington and 3 focal points. In ESA it's likely that the focal points will be Naomi, (Joseph) and Regis. We still have one to find in WA and ET - Asamoah has an idea. The main requirement is to be expert in survey design, implementation and M&E. Ditto with Ethiopia. Stan is leaving Dec. 1 and he won't be able to commit much time. I will find time with Stan to discuss who will attend the planning meetings, develop the evaluation design etc. around the priority questions.
• Funds will be flowing from IFPRI straight to these people? --> Yes, they will be IFPRI employed consultants.
• The M&E person needs to be 100% on the job --> It depends on the CVs we'll end up with.
• Is Mali covered? --> Yes, covered through the West Africa appointment.
• What qualifications / ToR do we need for the M&E consultants? --> Carlo will send draft ToR about this e.g. experience in agric, M&E, running and implementing surveys, ability to contract and outsource, data analysis expertise.
• Don't underestimate the backstopping you will need to provide. --> Yes, we're

Should we randomize? (Peter)

Districts are very heterogeneous. It is statistically desirable but is it practically possible?
Looking at practical design:

• We need to consider heterogeneity
• Some sites may allow designs that are not possible at others
• Analysis may allow us to compensate for practical shortcomings e.g. filtering pre-engagement sites
• We could do a phased roll-out of randomization
• Sample sizes (number of farms within village and villages)

Can we have a write-up on a process: timings / practicalities. e.g. this is where we really need to sit down and agree etc. (researchers and on-site + M&E team).

We should randomize when possible (controls can be randomized). This is all feasible in some way.
This needs continuous dialogue between on-site researchers and people drawing conclusions in and across sites.

It'd be good to have a small group to work on specific next steps - Sibiry, Peter, Dr. Karen Mancose - perhaps she could come and visit the other mega-sites. --> but this might clash with timing.